Deconstructing Alibaba Cloud's Preemptible Instance Pricing

Danielle Movsowitz Davidow; Orna Agmon Ben-Yehuda; Orr Dunkelman

doi:10.1145/3588195.3593001

Deconstructing Alibaba Cloud's Preemptible Instance Pricing

Danielle Movsowitz Davidow, Orna Agmon Ben-Yehuda, Orr Dunkelman

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Alibaba Cloud's preemptible instances are IaaS instances sold using an auction, similar to the spot instances offered by Amazon Web Services (AWS). Clients purchase them by bidding. The instances are terminated when supply is insufficient, or when the client's bid is lower than the "market price"- - which Alibaba Cloud publishes ever so often. We collected Alibaba Cloud's preemptible price traces from November 2018 to July 2021. This work analyzes the traces, and shows evidence that they experienced sharp lateral behavioral changes over time. We characterize the traces before and after these events. Our analysis highlights features in the traces that seem to be artificially engineered by an underlying pricing mechanism. Since engineered features may abruptly change, they are particularly interesting to clients planning their bidding strategy, or scholars who study them (e.g., to design new pricing mechanisms on the basis of real-life data).

Original language	English
Title of host publication	HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing
Publisher	Association for Computing Machinery, Inc
Pages	253-265
Number of pages	13
ISBN (Electronic)	9798400701559
DOIs	https://doi.org/10.1145/3588195.3593001
State	Published - 7 Aug 2023
Event	32nd International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2023 - Orlando, United States Duration: 16 Jun 2023 → 23 Jun 2023

Publication series

Name	HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing

Conference

Conference	32nd International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2023
Country/Territory	United States
City	Orlando
Period	16/06/23 → 23/06/23

Bibliographical note

Publisher Copyright:
© 2023 ACM.

Keywords

alibaba cloud
auctions
cloud computing
traces

ASJC Scopus subject areas

Information Systems
Software
Safety, Risk, Reliability and Quality
Artificial Intelligence
Computer Networks and Communications
Computer Science Applications
Hardware and Architecture

Access to Document

10.1145/3588195.3593001

Cite this

Movsowitz Davidow, D., Agmon Ben-Yehuda, O., & Dunkelman, O. (2023). Deconstructing Alibaba Cloud's Preemptible Instance Pricing. In HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing (pp. 253-265). (HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing). Association for Computing Machinery, Inc. https://doi.org/10.1145/3588195.3593001

Movsowitz Davidow, Danielle ; Agmon Ben-Yehuda, Orna ; Dunkelman, Orr. / Deconstructing Alibaba Cloud's Preemptible Instance Pricing. HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc, 2023. pp. 253-265 (HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing).

@inproceedings{90938fd93dd348a2a9b87d7d6459fbfb,

title = "Deconstructing Alibaba Cloud's Preemptible Instance Pricing",

abstract = "Alibaba Cloud's preemptible instances are IaaS instances sold using an auction, similar to the spot instances offered by Amazon Web Services (AWS). Clients purchase them by bidding. The instances are terminated when supply is insufficient, or when the client's bid is lower than the {"}market price{"}- - which Alibaba Cloud publishes ever so often. We collected Alibaba Cloud's preemptible price traces from November 2018 to July 2021. This work analyzes the traces, and shows evidence that they experienced sharp lateral behavioral changes over time. We characterize the traces before and after these events. Our analysis highlights features in the traces that seem to be artificially engineered by an underlying pricing mechanism. Since engineered features may abruptly change, they are particularly interesting to clients planning their bidding strategy, or scholars who study them (e.g., to design new pricing mechanisms on the basis of real-life data).",

keywords = "alibaba cloud, auctions, cloud computing, traces",

author = "{Movsowitz Davidow}, Danielle and {Agmon Ben-Yehuda}, Orna and Orr Dunkelman",

note = "Publisher Copyright: {\textcopyright} 2023 ACM.; 32nd International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2023 ; Conference date: 16-06-2023 Through 23-06-2023",

year = "2023",

month = aug,

day = "7",

doi = "10.1145/3588195.3593001",

language = "English",

series = "HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing",

publisher = "Association for Computing Machinery, Inc",

pages = "253--265",

booktitle = "HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing",

}

Movsowitz Davidow, D, Agmon Ben-Yehuda, O & Dunkelman, O 2023, Deconstructing Alibaba Cloud's Preemptible Instance Pricing. in HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing. HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing, Association for Computing Machinery, Inc, pp. 253-265, 32nd International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2023, Orlando, United States, 16/06/23. https://doi.org/10.1145/3588195.3593001

Deconstructing Alibaba Cloud's Preemptible Instance Pricing. / Movsowitz Davidow, Danielle; Agmon Ben-Yehuda, Orna; Dunkelman, Orr.
HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc, 2023. p. 253-265 (HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Deconstructing Alibaba Cloud's Preemptible Instance Pricing

AU - Movsowitz Davidow, Danielle

AU - Agmon Ben-Yehuda, Orna

AU - Dunkelman, Orr

PY - 2023/8/7

Y1 - 2023/8/7

N2 - Alibaba Cloud's preemptible instances are IaaS instances sold using an auction, similar to the spot instances offered by Amazon Web Services (AWS). Clients purchase them by bidding. The instances are terminated when supply is insufficient, or when the client's bid is lower than the "market price"- - which Alibaba Cloud publishes ever so often. We collected Alibaba Cloud's preemptible price traces from November 2018 to July 2021. This work analyzes the traces, and shows evidence that they experienced sharp lateral behavioral changes over time. We characterize the traces before and after these events. Our analysis highlights features in the traces that seem to be artificially engineered by an underlying pricing mechanism. Since engineered features may abruptly change, they are particularly interesting to clients planning their bidding strategy, or scholars who study them (e.g., to design new pricing mechanisms on the basis of real-life data).

AB - Alibaba Cloud's preemptible instances are IaaS instances sold using an auction, similar to the spot instances offered by Amazon Web Services (AWS). Clients purchase them by bidding. The instances are terminated when supply is insufficient, or when the client's bid is lower than the "market price"- - which Alibaba Cloud publishes ever so often. We collected Alibaba Cloud's preemptible price traces from November 2018 to July 2021. This work analyzes the traces, and shows evidence that they experienced sharp lateral behavioral changes over time. We characterize the traces before and after these events. Our analysis highlights features in the traces that seem to be artificially engineered by an underlying pricing mechanism. Since engineered features may abruptly change, they are particularly interesting to clients planning their bidding strategy, or scholars who study them (e.g., to design new pricing mechanisms on the basis of real-life data).

KW - alibaba cloud

KW - auctions

KW - cloud computing

KW - traces

UR - http://www.scopus.com/inward/record.url?scp=85169606013&partnerID=8YFLogxK

U2 - 10.1145/3588195.3593001

DO - 10.1145/3588195.3593001

M3 - Conference contribution

AN - SCOPUS:85169606013

T3 - HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing

SP - 253

EP - 265

BT - HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing

PB - Association for Computing Machinery, Inc

T2 - 32nd International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2023

Y2 - 16 June 2023 through 23 June 2023

ER -

Movsowitz Davidow D, Agmon Ben-Yehuda O, Dunkelman O. Deconstructing Alibaba Cloud's Preemptible Instance Pricing. In HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc. 2023. p. 253-265. (HPDC 2023 - Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing). doi: 10.1145/3588195.3593001

Deconstructing Alibaba Cloud's Preemptible Instance Pricing

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this